Mixing different IRs in an orchestral piece

BassClef

Senior Member
I'm still relatively new (hobbyist) at this VI game, and reverb is a certainly one of my many weak points.

I just watched the latest EastWest composer tip (video) on how to create a Fantasy track. It was great...learned a few things... enjoyed the piece. But I'm confused about his use of Spaces II reverb. He applied different IRs to the different stems... different spaces. I think he used 4 or 5 different ones. Now I can not argue with the outcome, as it sounded great, but this seems to contradict the whole idea of making the instruments all sound like they are in the same space. In my simple minded approach, I would just use one IR and vary the parameters like predelay and tail for various stems. Your thoughts?
 

Beat Kaufmann

Active Member
Without being able to see the video or hear the result, I cannot say much.

When it comes to placing instruments at different depths, you may be able to use different IRs. But then you should blend them all together with a "reverb over all" in one room to make it sound as natural as possible. That is also my experience.

But
A) in the end the result counts. And...
B) everybody is allowed to put videos on Youtube, even if the shown tips and tricks are not the best or even wrong... ;)

Beat
 
OP
BassClef

BassClef

Senior Member
Thanks for the reply Beat. In the East West video I referenced thats exactly what the composer did. After using different Spaces II IRs on various stems, he put a Lexicon simulation on the final mix bus.
 
OP
BassClef

BassClef

Senior Member
Beat... I really liked the video and especially the mixer layout... a few questions, if your don't mind.

NOTE: I currently use Logic's "summing" track stacks for my 7 orchestral sections, each sent to the bus with a reverb inserted, and then simply vary the send amount from each sum to the reverb... not very effective!

On turning off reverb tail:

1) Your chart shows a red "X" over the reverb tail on the "depth" convolution reverbs. Is this "X" (turning off the tail) accomplished by simply reducing the decay time for the reverb?
2) How do yo know how short? By ear or milliseconds?
3) And are all 4 "depths" set the same amount of tail reduction?

On creating the different depths with convolution reverb:
1) Do you do this by adjusting the dry/wet mix for each depth? And if so, do you do it by adjusting the "send amount" on each sum track? Or do you send 100% from sum to depth and adjust wet/dry inside of the reverb?
2) In addition to wet/dry adjustments for each depth, do you also use different pre-delay settings for each depth? (like 10, 20 30, 40 milliseconds or such)

On final algorithmic reverb:
1) On your final algorithmic reverb from the depth tracks, do you vary the amount of each of the 4 sends going to that final reverb, to try and have each "depth" with a slightly different length tail?

Thanks
 

Beat Kaufmann

Active Member
Beat... I really liked the video and especially the mixer layout... a few questions, if your don't mind.
...
On turning off reverb tail:
IRs often depict very natural spaces. If you want to try out my reverb concept, you have to go through all your IRs and find the ones that at 100% wet push the instruments furthest and at the same time sound less colored.
Once you have found such a "good" IR you have to fade it out so that only the first 100ms to 300ms (optimized via the ear) are used. That would be the ERs.
Now you will be able to use the dry/wet ratio to fade the instruments as much as you want. Of course, this works best with samples that were "dry recorded. If samples already have a room depth integrated, you can omit this step.

On creating the different depths with convolution reverb:
Please see read in the text above as well. If you want to acoustically simulate room depth, it usually means that sounds from a distance do not sound as brilliant. This can be supported by using an EQ and a hi-shelf to reduce the high frequencies at greater depths. But please don't ask for dB-values. You have to adjust everything until it sounds natural and to your taste.
Forget the predelay values. They remain at 0ms. You have chosen an IR before, which contains everything to create natural room depth. Because it is a mix ratio dry (0ms) and wet (IR) the right predelays are given by the IR automatically.
Further: It is often necessary to add effects to the instruments in the greater depths of the room so that they generate sufficient pressure again, which they have lost due to the distance and lack of brilliance. All this is set up and optimized separately for each depth in the corresponding BUS channel.

On final algorithmic reverb:
Attention Attention!!! ...each of the 4 sends going to that final reverb...
Forget everything with "SEND" in this concept.
You produce one BUS channel for each room depth. If you decide that the strings are to be provided with room depth 1, you loop the whole strings fully (100% 100%!!!) through this BUS 1. NO SEND!!!!

The Reverb-effect (Tail) that glues everything together is located in the output channel. About Tail: In a room there is only one reverb-tail and this sum of all echoes has everywhere about the same volume. What changes for a listener is the ratio "direct sound - reverb". So if an instrument is far away, the direct sound may be about as loud as all echoes and reflections (Tail)

In short: There is a reverb in the output channel where all BUS channels are summed up. This is where the "TAIL-Reverb" must be inserted. The dry/wet ratio is usually about 10%-25%, not more.
It's good if this effect doesn't contain ERs anymore, but only Tail - > Because this reverb effect should not add any more depth to the music.

And once again: The SEND function is never used in the whole concept!
----------------------------------------------------------------------------
About SEND
The possibility of SEND dates back to the very early years, when the studios had only one reverb unit (hardware) at their disposal. The SEND-function was very important so that all instruments could get some reverb.
Later, when the first computers were able to mix music, the reverb plug-ins were often so computationally intensive that after 1 or 2 reverb plug-ins the computer went down on its knees.
I remember the first times when I often had to have each audio track calculated for itself and then freeze it. And again, the SEND function was very welcome.

Today, when you can easily install 60-100 reverb instances without the computer breaking down completely, the send function doesn't make much sense anymore, especially when mixing orchestras. Maybe even more so in pop music, where it's all about a little reverb tail. In orchestral music, however, it's mostly about real room depths and other things. Some manufacturers even rely on each instrument having a reverb unit that adjusts the exact depth of the room including position (L<>R). e.g. Breeze2 with Precedence, but also MIR etc.

In summary: Nowadays, the "SEND" function is no longer necessary in conjunction with reverb and also often does not produce exactly what you want to achieve.
----------------------------------------------------------------------------

All the best
Beat